17:34:37 <SumitNaiksatam> #startmeeting networking_policy
17:34:38 <openstack> Meeting started Thu Apr 17 17:34:37 2014 UTC and is due to finish in 60 minutes.  The chair is SumitNaiksatam. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:34:39 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
17:34:41 <openstack> The meeting name has been set to 'networking_policy'
17:35:00 <SumitNaiksatam> #info New blueprint review process in effect
17:35:18 <SumitNaiksatam> most of you already know this, but just for the record
17:35:37 <SumitNaiksatam> #topic GP Blueprint
17:35:54 <SumitNaiksatam> so it seems like we will have to resubmit our blueprint into gerrit as well
17:36:14 <SumitNaiksatam> kind of frustrating since it has been reviewed over and over again over the past several months
17:36:27 <s3wong> SumitNaiksatam: Yes. You are going to do that? :-)
17:36:35 <SumitNaiksatam> hopefully we will have a quicker turnaround on this
17:36:55 <SumitNaiksatam> s3wong: i started collecting some of the material and stuffing it into the template
17:37:03 <SumitNaiksatam> s3wong: will run it by you guys
17:37:15 <s3wong> SumitNaiksatam: Sure. Thanks!
17:37:17 <SumitNaiksatam> s3wong:  or we can directly go to the gerrit review itself
17:37:43 <SumitNaiksatam> i am not sure about the level of detail that is expected in this new process
17:37:56 <banix> can others update the document once submitted to gerrit or it should be by whoever submits it
17:38:07 <SumitNaiksatam> i see that most blueprints have an average of about 250 lines
17:38:16 <SumitNaiksatam> banix: it should be possible to push a patch
17:38:22 <SumitNaiksatam> as in anyone should be able to do it
17:38:29 <banix> ok thx
17:38:33 <SumitNaiksatam> it will help to have some coordination thought
17:38:36 <SumitNaiksatam> *though
17:38:41 <banix> yup
17:38:48 <SumitNaiksatam> banix: this is just like another code patch
17:38:54 <SumitNaiksatam> mechanics remain the same
17:39:12 <SumitNaiksatam> banix: actually you already uploaded a patch, so i am preaching to the choir :-)
17:39:46 <SumitNaiksatam> i will try to get the first iteration out at the earliest
17:39:59 <banix> ok thanks
17:40:18 <SumitNaiksatam> i believe rkukura was going to suggest a change to the review template in terms of how the REST API details are captured
17:40:18 <s3wong> SumitNaiksatam: cool
17:40:28 <SumitNaiksatam> i thought i would wait for that
17:40:49 <SumitNaiksatam> else, documenting each and every API call for every resource would just blow up the document
17:41:09 <SumitNaiksatam> any more thoughts, questions, on this?
17:41:09 <rkukura> SumitNaiksatam: Is a table easy enough to do in RST format?
17:41:29 <SumitNaiksatam> rkukura: i was hoping to piggy back on the table you would create :-P
17:41:45 <SumitNaiksatam> rkukura: i was going to suggest that keep the columns wide enough
17:42:00 <SumitNaiksatam> rkukura: i dont think asciiflow will help there
17:42:18 <SumitNaiksatam> rkukura: perhaps there is a tool for ascii tables but i am not aware
17:42:21 <SumitNaiksatam> anyone know?
17:42:22 <thinrichs> There's definitely support for tables built into RST. Don't remember the details.
17:42:34 <rkukura> looking into it
17:42:37 <SumitNaiksatam> thinrichs: ah, thanks, guesses as much
17:43:04 <SumitNaiksatam> *guessed
17:43:46 <SumitNaiksatam> ok moving on
17:43:57 <rkukura> http://docutils.sourceforge.net/docs/user/rst/quickref.html#tables
17:44:07 <SumitNaiksatam> rkukura: great, thanks
17:44:47 <rkukura> SumitNaiksatam: I’ll propose a patch to the template with an example table.
17:45:09 <SumitNaiksatam> rkukura: great thanks, i was thinking if plain text might better for in place review
17:45:12 <SumitNaiksatam> not sure
17:45:50 <SumitNaiksatam> #topic Group Policy Wiki
17:45:58 <s3wong> Yes. If it isn't plain text, difficult to do gerrit review
17:46:55 <SumitNaiksatam> so far we were putting everything into the meeting wiki page
17:47:19 <banix> SumitNaiksatam: yeah need to clean that up and put references in a wiki; agree.
17:47:28 <SumitNaiksatam> since we have a lot more technical and planning content to deal with, a separate wiki page would be better
17:47:33 <SumitNaiksatam> #link https://wiki.openstack.org/wiki/Neutron/GroupPolicy
17:47:41 <SumitNaiksatam> banix: yeah, that was the thought
17:47:55 <SumitNaiksatam> banix: i havent yet moved anything to the new page
17:48:10 <SumitNaiksatam> just wanted to check if everyone was fine
17:48:12 <rkukura> SumitNaiksatam: not sure if you use emacs, but this could be helpful: http://docutils.sourceforge.net/docs/user/emacs.html#editing-tables-emacs-table-mode
17:48:22 <banix> SumitNaiksatam: sounds good; I can help
17:48:38 <SumitNaiksatam> rkukura: i am a vim guy, but thanks, will take a look
17:48:45 <SumitNaiksatam> banix: thanks, that would be great
17:49:13 <SumitNaiksatam> we can probably use the wiki for tracking progress as well
17:49:27 <SumitNaiksatam> which brings us to the progress update on the PoC
17:50:00 <SumitNaiksatam> since there is very little time left before the summit (and which we are targeting as a milestone), i would propose that we get tactical at this point
17:50:43 <SumitNaiksatam> so i wanted to bring the agenda down to track progress in each of the substreams for the PoC
17:50:49 <SumitNaiksatam> sound okay?
17:51:24 <SumitNaiksatam> everyone is quiet today
17:51:26 <banix> yes
17:51:31 <rkukura> SumitNaiksatam: +1
17:51:32 <s3wong> SumitNaiksatam: sounds good
17:51:42 <SumitNaiksatam> ah thanks guys :-)
17:51:53 <SumitNaiksatam> #topic Policy Model/Manager
17:51:55 <mandeep> SumitNaiksatam: Yes ;-)
17:52:32 <SumitNaiksatam> so i made some progress on the sumit/pm branch, but still behind on the shcedule
17:52:49 <SumitNaiksatam> i fixed a bunch of issues with the plugin to get the tests to work
17:53:17 <SumitNaiksatam> i was trying to make sure that the end-to-end scenario works
17:53:31 <rkukura> SumitNaiksatam: I added some stuff to this on my rkukura/mapping branch to test creating endpoint groups and get that working.
17:53:41 <SumitNaiksatam> i expected to much progress than that, but unfortunately thats where i am
17:53:45 <SumitNaiksatam> rkukura: yeah great thanks
17:54:00 <SumitNaiksatam> rkukura: i had not implemented those
17:54:31 <SumitNaiksatam> any questions for me (we will get to the policy driver shortly)
17:54:36 <banix> thnaks; any major departure from what we had discussed earlier?
17:54:39 <SumitNaiksatam> banix has also been reviewing
17:55:13 <SumitNaiksatam> we might need to have a discussion around the services part (again!)
17:55:28 <banix> yes wanted to look into services part
17:55:41 <SumitNaiksatam> but i had not reached anywhere close to that part of the implementation, so its not in the code anyways
17:56:04 <SumitNaiksatam> i got some feedback on the epg/contract relationship modeling as well
17:56:16 <SumitNaiksatam> so we might need to make some enhancements
17:56:39 <SumitNaiksatam> especially to handle your case banix where the EPG can be explicitly address (if required)
17:57:02 <SumitNaiksatam> *addressed
17:57:22 <s3wong> SumitNaiksatam: well, since this isn't the part about the enforcer yet - I am not updating. But currently I am assuming service doesn't get rendered until policy is passed - and still having to work with you and the team to finalize "service" object and how to derive that this means "firewall"
17:57:50 <SumitNaiksatam> s3wong: yes, sorry that this is part is not crystallized yet
17:58:04 <SumitNaiksatam> s3wong: i think all of us see the difficulty in wrapping our heads around this
17:58:41 <s3wong> SumitNaiksatam: is OK - plenty of stuff to look at in the plugin and how it can invoke firewall object before having to code the northbound piece
17:58:43 <SumitNaiksatam> s3wong: part of the problem is that the services’ framework in neutron itself is not as much baked
17:59:12 <SumitNaiksatam> s3wong: we all have been spending time independently on that, but we know where that is ;-)
17:59:34 <SumitNaiksatam> more white board discussions required i guess
17:59:43 <s3wong> SumitNaiksatam: absolutely
18:00:00 <banix> there was a mention of changes in policy context yesterday; is that what you thought we needed in the context of policies?
18:00:53 <SumitNaiksatam> banix: yes, one on the services, the second along the lines of the change to the EPG/Contract relationship i was mentioning earlier
18:01:40 <banix> is this a good time to discuss that?
18:02:36 <SumitNaiksatam> banix: perhaps having a modified diagram would be helpful
18:02:49 <banix> SumitNaiksatam: yes that would help
18:02:56 <mandeep> SumitNaiksatam: Is this done best in context of the new bluprint ?
18:02:58 <SumitNaiksatam> banix: i got caught up with doing the blueprint and did not get a chance to update the diagram
18:03:04 <SumitNaiksatam> mandeep: yes
18:03:38 <banix> SumitNaiksatam: sure; we can discuss later
18:03:57 <SumitNaiksatam> banix: yeah, its probably a longer discussion
18:04:04 <SumitNaiksatam> lets get through the status first
18:04:20 <SumitNaiksatam> #topic Policy plugin driver update
18:04:35 <SumitNaiksatam> rkukura: i believe you made good progress on this
18:04:40 <rkukura> I’m working towards a short term goal:
18:04:52 <SergeyLukjanov> hey folks, you're out of time
18:05:03 <rkukura> To drive creation of the neutron network, subnet, and port resources via the group API, with all traffic allowed
18:05:13 <SumitNaiksatam> SergeyLukjanov: bummer!
18:05:24 <SumitNaiksatam> SergeyLukjanov: we thought we had another 30 mins
18:05:36 <SergeyLukjanov> nope :)
18:05:36 <SumitNaiksatam> rkukura: sorry, hang on
18:05:40 <rkukura> Then s3wong and I can work from that on mapping the policies
18:06:06 <SumitNaiksatam> SergeyLukjanov: okay sure, we can end this meeting here
18:06:15 <SergeyLukjanov> SumitNaiksatam, thx
18:06:15 <SergeyLukjanov> sahara folks, ping
18:06:21 <s3wong> SumitNaiksatam: Is #openstack-meeting-3 available?
18:06:26 <SumitNaiksatam> group policy folks, lets join on -meeting-3
18:06:29 <SumitNaiksatam> s3wong: yeah
18:06:33 <SumitNaiksatam> #endmeeting